JPH1155664A - Binary shape signal encoding device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enhance the resolution of a decoded image gradually by simultaneously encoding video (enhancement layer) which has high resolution and video (base layer) which has low resolution and encoding the enhancement layer, based on the base layer. SOLUTION: A 1st enhancement layer encoding part 230 encodes a vertical enhancement layer, based on a 1st sample block on a line L13, a 2nd sample block on a line L16 and a vertical insertion signal on a line L17. A 2nd enhancement layer encoding part 245 encodes a horizontal enhancement layer based on a 1st block on a line L11, a 2nd block on a line L12 and a horizontal insertion signal on a line L20. An MUX 250 multiplexes data of an encoding base layer, data of an encoding horizontal enhancement layer and data of an encoding vertical enhancement layer and outputs them to a decoder.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a binary shape signal encoding apparatus, and more particularly, to a binary shape signal encoding apparatus which realizes scalability for gradually increasing the resolution of an image.

[0002]

2. Description of the Related Art Generally, in a digital video system such as a videophone or an electronic conference, a video frame signal is composed of a series of digital data called "pixels". Of digital data is required.

[0003] However, the available frequency bandwidth over a normal transmission channel is limited, so transmitting large amounts of digital data over that channel is especially necessary for low bit rates such as video telephony and teleconferencing. For a rate video signal encoder, the amount of data to be transmitted must be compressed or reduced using various data compression techniques.

In a low bit rate video signal coding system, one of the methods for coding a video signal is a so-called object-oriented analysis / synthesis coding method (Object-Oriented An).
analysis-Synthesis coding technique). According to this object-oriented analysis / synthesis coding technique, an input video signal is divided into a plurality of objects (objects), and codes comprising three sets of parameters that define the motion, outline, and pixel data of each object are different. Is handled through the licensing channel.

As an example of such an object-oriented analysis / synthesis encoding method, a so-called MPEG-4 (Moving Picture Exposure) is used.
Although there is an erts Group phase 4), this MPEG-4 is content-based interactivity in applications such as low bit rate communication, interactive multimedia (eg, games, interactive television, etc.) and area monitoring equipment. Provide an audiovisual coding standard that enables improved coding efficiency and / or universal accessibility.

In MPEG-4, an input video image is divided into a plurality of video object planes (VOPs) corresponding to entities in a bitstream that are accessible and operable by a user. VOP
Is also defined as an object and may be represented by a square whose width and height are the smallest multiple of 16 pixels (the size of a macroblock) surrounding each object. Therefore, the encoding unit uses VOP
The input video image is handled in units, that is, in units of objects.

[0007] A VOP described in MPEG-4 is composed of luminance data, and has shape information represented by a binary mask and color information composed of chrominance data. In the binary mask, one binary (eg, 0) represents a pixel located outside the object (background pixel) in the VOP, and the other binary (eg, 1) represents a pixel located inside the object (object pixel). ) Respectively. Binary shape information indicating the position and shape of an object is represented by a BAB (BAB; Binary alpha block) encoded by a known bitmap-based shape encoding technique.

For example, when a bitmap-based shape coding technique such as a content-based arithmetic coding technique is used, coding error data and a binary shape signal in the current frame (or VOP) are obtained by motion estimation and motion compensation. When motion compensation information representing the difference between the binary shape signal in the previous frame (or VOP) and the most similar binary motion signal is obtained, the encoded binary value representing the shape and position of the object in the intra mode is obtained. Shape information can be obtained.

However, the encoded binary shape information obtained as described above is decoded by the decoder to generate a reconstructed video having only a predetermined resolution. Therefore, when a video having a higher resolution is required, the binary shape information is usually encoded through scalability that gradually increases the resolution of the decoded video. That is, an image having a lower resolution (base layer) is encoded, and additional information is added based on the base layer to encode an image (enhancement layer) having a higher resolution.

FIG. 1 is a block diagram showing a conventional binary signal encoding apparatus 100 using a bitmap-based encoding technique. This binary shape signal encoding device 10
0 is the sub-sampling unit 110 and the base layer encoding unit 12
0, an enhancement layer coding unit 130 and a multiplexer (MUX) 140. Further, the enhancement layer coding section 130 includes an enhancement layer reconstructing section 131, a subtraction section 133, and an island coding section 135.

A BAB composed of binary pixels representing an object and a background in a frame or VOP (for example, a block composed of 16 × 16 binary pixels) is supplied to a sub-sampling unit 110 and an enhancement layer encoding unit 130 through a line L1. Are respectively input to the subtraction unit 133. The sub-sampling unit 110 receives the 16 × 16 pixel BA
B is sampled, and a sample block including 8 × 8 pixels is supplied to the base layer encoding unit 120 and the enhancement layer reconstructing unit 131 via the line L2.
The sampling method of the sub-sampling unit 110 is as follows. First, a BAB of 16 × 16 pixels is divided into blocks of 2 × 2 pixels, and binary pixel values frequently appearing in each block of 2 × 2 pixels are converted to 2 × 2 pixels. Select as a representative value for a block of pixels. At this time, when the number of two binary pixel values in the block of 2 × 2 pixels is the same, the binary pixel values are selected according to a predetermined rule. Thereafter, all selected binaries are corrected in order to form a sample block of 8 × 8 pixels.

The base layer encoding unit 120 encodes the received sub-block using an encoding method such as a bitmap-based shape encoding method, and supplies the encoded base layer data to the MUX 140 via the line L3.

The enhancement layer reconstructing unit 131 reconstructs the enhancement layer based on the sub-block according to a predetermined rule, and supplies the reconstructed enhancement layer to the subtraction unit 133 via the line L5. The encoded enhancement layer is encoded, and the encoded enhancement layer data is supplied to the MUX 140 via the line L4. Here, according to a predetermined rule, a block of 2 × 2 pixels constituting four binary pixels is reconstructed for each binary pixel in the sub-block. Here, all four pixels in the reconstructed block have the same pixel value as each binary pixel in the sample block. Thereafter, the BABs of 2 × 2 pixels are combined in order to form a 16 × 16 pixel BAB. This BA
B is supplied in order as an enhancement layer. In the decoder, since the reconstruction method used in the enhancement layer encoding unit 130 is known, the enhancement layer is reconstructed using the method.

Next, the subtraction unit 133 subtracts the reconstructed enhancement layer input through the line L5 from the BAB input through the line L1, and outputs an error data block composed of 16 × 16 pixels as a result of the subtraction. It is supplied to the island encoding unit 135. Here, the error data block has a first binary pixel value (eg, 1) and a second binary pixel value (eg, 0), ie, the first binary pixel value is reconstructed as BAB. The second binary pixel value represents a pixel whose subtraction result is non-zero, and a second binary pixel value represents a pixel whose subtraction result is zero.

The island encoding unit 135 is, for example, RC
Using B (Reference Contour Based) coding technique,
The error data block from the subtraction unit 133 is island-encoded, and the island-encoded data is supplied to the MUX 140 via the line L6.

The MUX 140 multiplexes the coded base layer data on the line L3, the coded enhancement layer data on the line L4, and the island coded data on the line L6, and sends the multiplexed data to the decoder at the receiving end. Supply the transmitter for transmission.

The data embodied in consideration of the scalability can be decoded by a decoder using various methods. One of the methods is to decode only a lower coding layer such as a coding base layer to reconstruct a video having a lower resolution. However, it is also possible to decode the base layer and some higher layers to increase the resolution of the video. In order to further increase the resolution, regardless of whether higher layers can be decoded or not, assuming that the lower layers are decoded before the higher layers, all transmitted layers are decoded and the original video is decoded. An image having the same resolution as the above can be obtained. The encoding / decoding method embodied in consideration of scalability can reduce errors, prevent bit loss, and transmit a video signal with high resolution.

However, in the conventional enhancement layer encoding apparatus, the correlation between the reconstructed enhancement layer obtained from the BAB and the BAB itself is not taken into account, without sufficiently considering the correlation between consecutive binary pixels. Many errors will occur. As a result, the transmission of island coded data requires a large number of bits, and furthermore, in the inter mode, the coding efficiency is not increased by using the correlation between the current VOP and the previous VOP. Therefore, it is difficult for the conventional enhancement layer coding apparatus to effectively code the enhancement layer.

[0019]

SUMMARY OF THE INVENTION Accordingly, it is a primary object of the present invention to provide a binary signal encoding apparatus which realizes scalability for increasing the resolution of a decoded video step by step.

[0020]

In order to achieve the above object, according to the present invention, M × N provided in a video signal is provided.
(M and N are each an even positive integer)
y) A binary shape signal encoding device for encoding a binary alpha block (BAB) composed of pixels, wherein each horizontal line of the BAB is sampled and the BAB is sampled.
Horizontal sub-sampling means for generating a first block starting from a first horizontal line located at the top of the first block or a second horizontal line located next to the top, and sampling each vertical line of the first block; Vertical subsampling means for generating, as a base layer, a first sample block starting from a first vertical line which is a leftmost vertical line of the first block or a second vertical line located next to the leftmost line, Encoding a first sample block to generate encoded base layer data, and encoding an enhancement layer based on the BAB, the first block and the first sample block, Enhancement for generating encoded horizontal enhancement layer data and encoded vertical enhancement layer data Is the binary shape signal coding apparatus characterized is provided to include a layer coding unit.

[0021]

Preferred embodiments of the present invention will be described below in detail with reference to the drawings. In accordance with the present invention, scalability is used to code a high resolution video (eg, an enhancement layer) at the same time as a low resolution video (eg, a base layer), wherein the enhancement layer is coded based on the base layer. You.

FIG. 2 is a block diagram showing a binary signal encoding apparatus 200 according to the present invention. The binary shape signal encoding device 200 includes a horizontal sub-sampling unit 205, a vertical sub-sampling unit 210, a base layer encoding unit 215, and an enhancement layer encoding unit 290.
And MUX250. The enhancement layer coding unit 290 includes two subtraction units 220 and 225, and first and second enhancement layer coding units 230 and 24.
5, first and second line supply units 235 and 240. Binary shape information, such as a video frame or video object plane (VOP), consisting of a plurality of binary alpha blocks BAB, each of M × N binary pixels, represents objects and backgrounds in the video frame or VOP. , BAB unit for horizontal sub-sampling unit 2 through line L10
05 and the subtraction unit 225. Here, M and N are each even numbers. For convenience of explanation, it is assumed that M and N are 16 each. FIG. 5 shows a B composed of 16 × 16 pixels.
AB501 is shown, in which black parts are binary pixels representing the object and white parts represent the background.

The horizontal subsampling unit 205 samples each horizontal line of the received BAB,
One of AB's first horizontal line or second horizontal line
Generate a first block starting with The first horizontal line is the uppermost line of BAB. For convenience of explanation, the horizontal sub-sampling unit 205 samples eight even-numbered horizontal lines consecutive in the BAB, and
A first block composed of x16 pixels is vertically directed through a line L11 to a sub-sampling unit 210, a subtracting unit 220
Second enhancement layer encoding section 245 and subtraction section 22
5 respectively. FIG. 6 is a schematic diagram of a first block composed of 8 × 16 pixels, where a black portion represents an object and a white portion represents a background. The same applies to black portions and white portions in FIGS. 7 to 9.

The vertical sub-sampling section 210 samples each vertical line of the first block received from the horizontal sub-sampling section 205, and outputs the resulting first sample block to the first vertical line or the second vertical line of the BAB. Occurs as a base layer starting from one of the vertical lines. The first vertical line is the leftmost line in the first block. For convenience of explanation,
The vertical sub-sampling unit 210 receives the received first
Eight even-numbered vertical lines that are continuous in the block are sampled, and a base layer encoding unit 215, a subtraction unit 220, and a first enhancement are set through a line L13 using a first sample block including 8 × 8 pixels as a base layer. Each is supplied to the layer encoding unit 230. In FIG.
A schematic diagram of a first sample block 505 composed of 8 × 8 pixels transmitted through a line L13 is shown.

In the case of the intra mode, the base layer coding unit
215 encodes the received base layer (ie, the first sample block) using a conventional encoding method such as a bitmap-based shape encoding technique, and supplies the encoded base layer data to the MUX 250. On the other hand, in the case of the inter mode, the motion prediction and motion compensation unit (not shown) incorporated in the base layer encoding unit 215 includes the first sample block of the current frame and the base layer encoding unit. Compare with the corresponding block of the previous frame stored in the current frame memory embedded in H.215. As a result, the first sample block of the previous frame that is most similar to the first sample block of the current frame is selected as the estimated first sample block.

Next, motion vector information expressed as a two-dimensional vector composed of horizontal and vertical components is obtained. This motion vector information represents the displacement between the first sample block in the current frame and the estimated first sample block in the previous frame. The difference between the first sample block of the current frame and the estimated first sample block of the previous frame is encoded to generate encoding error data. Thereafter, the motion vector information and the encoded error data are combined as encoded base layer data and input to the MUX 250 via the line L14.

A base layer reconstructing unit (not shown) reconstructs a first sample block of the current frame based on the coding error data and the predicted first sample block. Next, the reconstructed first sample block is:
The data is stored at a set position in the first frame memory so as to be used for motion estimation and motion compensation of a subsequent frame. At the same time, the motion vector information is transmitted to the first line supply unit 23 in the enhancement layer encoding unit 290 through the line L15.
5 and the second line supply unit 240.

In the case of the intra mode, the enhancement layer coding unit 290 includes a BAB on the line L10, a first block on the line L11, and a first block on the line L13.
The enhancement layer is encoded based on the sample block, and the encoded vertical enhancement layer data is supplied to the MUX 250 via a line L18 and the encoded horizontal enhancement layer data is supplied to the MUX 250 via a line L21.

On the other hand, in the case of the inter mode, the enhancement layer encoding unit 290 also encodes the motion vector information input through the line L15 in addition to the case of the above-described intra mode, and transmits the information through the line L18. The data of the coded vertical enhancement layer is converted to the MUX of the coded horizontal enhancement layer data via a line L21.
250 respectively. Hereinafter, the operation of the enhancement layer coding unit 290 will be described in detail.

The subtraction unit 220 subtracts the first sample block on the line L13 from the first block on the line L11, and outputs the resulting second sample block via the line L16 to the first enhancement layer coding unit 230.
And to the first line supply unit 235. The second sample block includes eight odd-numbered vertical lines in the first block. On the other hand, the subtraction unit 225 similarly subtracts the first block on the line L11 from the BAB on the line L10, and outputs the resulting second block via the line L12 to the second enhancement layer coding unit 245 and the second block. Each is supplied to the line supply unit 240. 8
The second block composed of × 16 pixels is composed of eight odd-numbered horizontal lines in the BAB composed of 16 × 16 pixels.

In the case of the intra mode, the first line supply unit 235 stores only a second sample block of 8 × 8 pixels at a corresponding position in a second frame memory (not shown). In the case of the inter mode, the first line supply unit 23
5 stores a second sample block of 8 × 8 pixels in the previous frame stored in the second frame memory based on the motion vector information input through the line L15 in addition to storing the second sample block. Among them, a first extraction block that satisfies a predetermined first threshold is extracted. Here, the predetermined first threshold value is such that the horizontal and vertical distance between the second sample block composed of 8 × 8 pixels on the line L16 and the first extraction block are the horizontal and vertical components of the motion vector. Is set to
After that, the first extraction block is supplied to the first enhancement layer coding unit 230 as a set of vertical insertion lines via a line L17.

In the case of the intra mode, the second line supply unit 240 stores only the second block composed of 8 × 16 pixels at a corresponding position in the third frame memory (not shown). In the case of the inter mode, the second line supply unit 240
In addition to storing the second block, based on the motion vector information input through the line L15, among the second blocks of 8 × 16 pixels in the previous frame stored in the third frame memory, The second extraction block that satisfies the predetermined second threshold is extracted. here,
The predetermined second threshold value is 8 × 16 on the line L12.
The horizontal and vertical distance between the second block of pixels and the second extraction block is twice the horizontal component of the motion vector,
It is set to be one time the vertical component. afterwards,
The second extraction block is provided as a set of horizontal insertion lines through a line L20 as a second enhancement layer encoder 24.
5 is supplied.

According to the preferred embodiment of the present invention, in the case of the intra mode, the first enhancement layer coding unit 230 performs the second sample block on the line L16 and the line L1
3, the vertical enhancement layer is coded based on the first sample block, and the resulting coded vertical enhancement layer data is supplied to the MUX 250. In the case of the inter mode, the first enhancement layer coding unit 230 performs the coding of the vertical enhancement layer based on the set of vertical insertion lines input through the line L17 in addition to the case of the intra mode. Then, the resulting encoded vertical enhancement layer data is supplied to MUX 250 via line L18.

According to the first preferred embodiment of the present invention, in the case of the intra mode, the second enhancement layer encoder 24
Reference numeral 5 performs encoding of the horizontal enhancement layer based on the first block on the line L11 and the second block on the line L12, and supplies the resulting encoded data of the encoded horizontal enhancement layer to the MUX 250. Further, in the case of the inter mode, the second enhancement layer coding unit 245 performs coding of the horizontal enhancement layer based on a set of horizontal insertion lines input through the line L20 in addition to the case of the intra mode. Then, the resulting data of the encoded horizontal enhancement layer is supplied to the MUX 250 via the line L21.

The MUX 250 multiplexes the coded base layer data from the line L14, the coded horizontal enhancement layer data from the line L18, and the coded vertical enhancement layer data from the line L21. The encoded data is transmitted to a transmitter (not shown) so as to be transmitted to a decoder (not shown) at the receiving end.

FIG. 3 is a detailed block diagram of the first enhancement layer coding unit 230 in FIG. 2, and its operation will be described in detail.

First enhancement layer coding section 230
Comprises an interpolation unit 300, a first enhancement bit calculation unit 310, a second enhancement bit calculation unit 330, a bit comparison unit 350, and an encoding enhancement layer data selection unit 360. The interpolation unit 300 is 8 × 8
Using a set of vertical lines (for example, the set of vertical lines 507 shown in FIG. 15) obtained from a block of pixels (for example, the first sample block 505 shown in FIG. 7), a predetermined vertical interpolation rule or The method configures a set of vertical interpolation lines (eg, a set of vertical interpolation lines 509 illustrated in FIG. 16). The set 509 of vertical interpolation lines is supplied to the first enhancement bit calculation unit 310 via the line L30.

A predetermined vertical interpolation method for forming a set of vertical interpolation lines is also known in a decoder. Thus, the decoder decodes the received encoded base layer to generate a decoded base layer. Thereafter, the decoder generates a set of vertical interpolation lines identical to the set of vertical interpolation lines formed by the interpolation unit 300 by using a predetermined vertical interpolation method based on the decoding base layer. This predetermined vertical interpolation method can be variously implemented according to the preferred embodiment of the present invention. The vertical interpolation method according to the preferred embodiment of the present invention will be described below with reference to FIGS.

According to the vertical interpolation method of the present invention, a sample block (ie, a first sample block of 8 × 8 pixels) is up-sampled into an interpolation block having the same size as a sub-block by the vertical interpolation process. .

FIGS. 5 to 7 are schematic diagrams showing sub-blocks generated by the sampling process and corresponding sample blocks, respectively. The sub-block (ie, FIG.
The binary alpha block 501) of 16 × 16 pixels shown in FIG. 7 is subjected to the sampling process performed by the horizontal sub-sampling unit 205 and the vertical sub-sampling unit 210, and the sample block of 8 × 8 pixels shown in FIG. 505. 6 and 7, black portions represent object pixels constituting the object, and white portions represent background pixels.

In the vertical interpolation process, in order to generate an interpolation block, a sample block is first divided vertically to form a plurality of reference lines (ie, vertical lines of the first sample block). Thereafter, a line segment and a non-line segment are detected on each reference line. Here, each line segment is 1
Represented by one or more consecutive object pixels, each non-linear segment is defined by one or more consecutive background pixels. An interpolation line (that is, a vertical interpolation line) is formed based on these two line segments on the reference line, and forms an interpolation block together with the reference line.

Referring to FIG. 6, there is shown a schematic diagram for explaining a predetermined method of forming a corresponding interpolation line based on a reference line of a sample block. According to the invention, 1
One interpolation line is formed based on two adjacent reference lines.

First, referring to FIGS. 11 to 13, there is a schematic diagram showing a case where two reference lines have the same segment number.

As shown in FIG. 11, when there are overlapping line segments 10A and 20A on two adjacent reference lines 10 and 20, respectively, a line segment 30A on the interpolation line 30 generated by the two reference lines 10 and 20. Is determined based on the positions of the start point and the end point of the two overlapping line segments 10A and 20A. Therefore, the start point 2 of the line segment 10A and the start point 1 of the line segment 20A are averaged, and the resulting average value 1.5 is rounded down to 1 to determine the start point of the line segment 30A. Similarly, overlapping line segment 10A
Average value between the end point 5 of the line segment 5 and the end point 6 of the overlap line segment 20A.
5 is rounded down to 5, and determined as the end point of the line segment 30A. In the above description, when both reference lines 10 and 20 overlap each other, both overlapping line segments 10A and 20A overlap. The overlap between the reference line 10 and the reference line 20 is performed by comparing the positions of the object pixels included in each line segment on both the reference lines 10 and 20.

On the other hand, when there are both non-overlapping line segments 40A and 50A on both reference lines 40 and 50, the non-overlapping line segments 40A and 50A
As shown in FIG. 12, each of A has one pixel at the first pixel position or the last pixel position of the corresponding reference line. Here, the non-overlapping line segments 40A and 50A are both reference lines 40 and 50.
Represents the non-overlapping line segments when overlapping at the top of each, line segment 60A on interpolation line 60 is the number of pixels in non-overlapping line segment 40A, and line segment 60B is the number of pixels in non-overlapping line segment 50A. Are respectively generated based on That is, the line segment 60A has two object pixels that are half the number of object pixels on the non-overlapping line segment 40A, and starts from the first pixel position of the interpolation line 60. The line segment 60B has one object pixel at the last pixel position of the interpolation line 60, and the last pixel position is
It is determined by dividing the number of object pixels on A by 2 and truncating the resulting value.

When there are non-overlapping line segments 70A and 80A on both reference lines 70 and 80, the line segments are sequentially located as shown in FIG. Line segment 9 on interpolation line 90
0A is determined based on the start point of the non-overlapping line segment 70A, and the line segment 90B is determined based on the end point of the non-overlapping line segment 80A. Line segment 9
The start point of 0A and the end point of line segment 90B are calculated as in the following equation (1).

[0047]

SP = (3 * P + 1 * Q) / 4 EP = (1 * P + 3 * Q) / 4

Here, SP is the start point of the line segment on the interpolation line,
EP represents the end point of the line segment on the interpolation line, P represents the start point of the line segment on the reference line corresponding to the interpolation line, and Q represents the end point of the line segment on the reference line corresponding to the interpolation line. If the calculated value on the right side of the equation is not an integer, SP or EP is obtained by truncating the calculated value.

FIG. 14 is a schematic diagram showing a case where two reference lines have different numbers of line segments. The first reference line 15 is composed of one line segment, and the second reference line 25 is composed of three line segments. Is made up of minutes. In this case, the interpolation line 35 is
And the reference line 25.
Thus, the interpolation line 35 has object pixels that are commonly included in the line segments of the reference lines 15 and 25. In FIG. 14, all pixels on the reference line 15 are object pixels, so that the interpolation line 35 has the same pixel pattern as the reference line 25.

The predetermined method described with reference to FIG. 6 can be applied to both vertical and horizontal interpolation processes. Hereinafter, it will be described how the predetermined method is actually employed in the vertical interpolation process, and the upsampling process of the sample block will be described in detail.

Referring to FIG. 7, a schematic diagram illustrating the vertical interpolation process according to the preferred embodiment of the present invention is shown. First, the sample blocks 505 shown in FIG. 7 have indexes V2, V4,
.., V16 are assigned to eight vertical reference lines. Each of the eight vertical reference lines has a series of eight pixels. Once each of the vertical reference lines V2 to V16 is determined, the number of line segments on each of the vertical reference lines V2 to V16 starts from the leftmost vertical reference line V2, and is 2, 1, 1,.
1, 3, 1, 2, and 2. By comparing the number of line segments on each vertical reference line with the number of line segments on the adjacent vertical reference line, a vertical interpolation line (for example, FIG.
V1, V3,..., V15) are formed by the predetermined method described with reference to FIG.
To V16. As can be seen from each index, each vertical reference line is represented by an (i + 1) -th position index, and each vertical interpolation line is represented by an i-th position index. Here, i is an odd number 1, 3,... Therefore, for example, each vertical interpolation line V3 to V15 has two adjacent vertical reference lines (for example, V2 and V4, V4 and V6,..., V
14 and V16). However, since only the vertical reference line V2 corresponds to the vertical interpolation line V1, the vertical interpolation line V1 is determined by copying the vertical reference line V2.

In the above, each index assigned to each vertical interpolation line and vertical reference line may be differently assigned. That is, odd indexes V1 to V15 may be assigned to each vertical reference line, and odd indexes V2 to V16 may be assigned to each vertical interpolation line.

The vertical interpolation lines V1 to V15 form a set of the vertical interpolation lines 509 shown in FIG.
And at the same time, is supplied to the first enhancement bit calculator 310.

First enhancement bit calculation section 310
Receives a set 509 of vertical interpolation lines input from the interpolation unit 300 via the line L30 and a second sample block composed of 8 × 8 pixels input from the subtraction unit 220 via the line L16. The first enhancement bit calculator 310 receives the set 509 of the received vertical interpolation line and the second
The first enhancement layer is coded based on the sample block, and the resulting data of the first coded enhancement layer is supplied to the bit comparison unit 350 via the line L33. Thereafter, the first enhancement bit calculator 310 calculates the bits of the first encoded vertical enhancement layer data, and supplies the calculation result to the encoded enhancement layer data selector 360 as the first vertical bit data via the line L34. In FIG.
FIG. 4 is a detailed block diagram of the first enhancement bit calculator 310 in FIG.

First enhancement bit calculation section 310
Are the first line analyzer 410 and the second line analyzer 41
5, line segment comparing section 420, line selecting section 440, first line adjusting section 455, second line adjusting section 460, shape reconstructing section 465, subtracting section 430, island encoding section 435,
It comprises a mode encoding unit 470 and a data formatting and bit calculation unit 480.

The first line analysis unit 410 determines whether the line L30
Is analyzed to find the number of line segments in each vertical interpolation line, the length of each line segment, the start point and the end point, and divide the search result into a first sample block line. It is supplied to the line segment comparison unit 420 as analysis information. Similarly, the second line analysis unit 415 analyzes the set of vertical lines of the second sample block including 8 × 8 pixels input through the line L16, and determines the number of line segments in each vertical line and the number of line segments in each vertical line. And the search result is supplied to the line segment comparison unit 420 as second sample block line analysis information.

The line segment comparing section 420 first stores the received first and second sample block line analysis information in a built-in memory (not shown), and then stores the first and second sample block line analysis information. Based on the information, the number of line segments in each vertical interpolation line is compared with the number of line segments of each corresponding vertical line in the second sample block. here,
Each corresponding vertical line in the second sample block is a vertical line at the same position as each vertical interpolation line. If the number of line segments in two corresponding lines (ie, each vertical interpolation line and each corresponding vertical line) is the same, the first line analysis information for the vertical interpolation line and each vertical line in the second sample block Is supplied to the first line adjustment unit 455 through the line L43.

On the other hand, if the number of line segments in each vertical interpolation line and each corresponding vertical line is not the same, the line comparison unit 4
20 generates third and fourth line analysis information for the vertical lines of the second sample block located on the left and right sides of the vertical interpolation line, respectively.

However, if the vertical interpolation line is the leftmost vertical line in the set of vertical interpolation lines, the line segment comparison unit 42
0 uses the line analysis information for the rightmost vertical line in the second sample block stored in the memory in advance instead of the third line analysis information for the leftmost vertical line in the set of vertical interpolation lines. As the third line analysis information for Then, the first, second, and third line analysis information extracted from the memory is sent to the line selection unit 440 through the line L41.

The first line adjustment unit 455 adjusts the length of the line segment in each vertical interpolation line based on the first and second line analysis information so as to be the same as the corresponding vertical line in the second sample block. By doing so, first length adjustment information indicating the line adjustment state is generated. For example, the first
The line adjustment unit 445 may change the start point and the end point of the line segment in each vertical interpolation line.

The first line adjustment unit 445 sends a first length adjustment mode signal indicating whether or not the length of the line segment has been adjusted, together with the first length adjustment information, through the line L46 to the shape reconstruction unit 465. , A mode encoding unit 470 and a data formatting and bit calculation unit 480.

The line selection section 440 includes the first, third and fourth lines.
Based on the line analysis information, the number of line segments in each vertical interpolation line is compared with the number of vertical lines located on the right and left sides of each vertical interpolation line.
0 selects one of the left and right vertical lines according to a predetermined selection rule.

After that, the line selection section 440 transmits the line analysis information for the selected vertical line and a first line selection mode signal for notifying that the vertical line has been selected based on a predetermined rule through the line L44. The data is sent to the shape reconstruction unit 465, the mode encoding unit 470, and the data formatting and bit calculation unit 480, respectively.

However, when the number of line segments in each vertical interpolation line is the same as the number of vertical lines located on the right and left sides of each vertical interpolation line, the line selection unit 440 sets the same line as the vertical interpolation line. Select a vertical line with a fraction.

Next, the line selection section 440 sends line analysis information for the selected vertical line and a second line selection mode signal for notifying whether the vertical line has been selected between the right vertical line and the left vertical line, to the line L42. Is supplied to the second line adjustment unit 460 via the. Further, the line selection unit 440 also sends the first line analysis information to the second line adjustment unit 46.
Supply 0.

The second line adjustment unit 460 uses the same method as the line adjustment unit 455 to determine the length of each vertical interpolation line based on the line analysis information for the vertical line, the first line selection mode signal, and the first line analysis information. And the resulting second length adjustment information, a second length adjustment mode signal indicating whether or not the length adjustment has been performed, and a second line selection mode signal are transmitted through the line L45 to the shape reconstruction unit 465. , A mode encoding unit 470 and a data formatting and bit calculation unit 480.

The shape reconstruction unit 465 includes line analysis information and a first line selection mode signal for the selected vertical line input through the line L44, second length adjustment information input through the line L45, and second length adjustment information. An adjustment mode signal, a second line selection mode signal, and a line L46.
A second sample block composed of 8 × 8 pixels is reconstructed based on the first length adjustment information and the first length adjustment mode signal input through to form a reconstructed second sample block. Thereafter, the reconstructed second sample block is supplied to the subtraction unit 430.

The subtraction section 430 calculates the second
The reconstructed second sample block from the shape reconstruction unit 465 is subtracted from the sample block, and the difference is supplied to the island encoding unit 435 as an error data block including 8 × 8 pixels. The error data block is composed of a first binary (eg, 1) and a second binary (eg, 0), where the first binary is the difference result between the BAB and the reconstructed enhancement layer block. Represents a non-zero pixel, and the second binary represents a pixel for which the difference result between the BAB and the reconstructed enhancement layer block is zero.

FIG. 8 shows a reconstructed second sample block 511 formed by the shape reconstruction unit 465 using the vertical interpolation line set 509 from the line analysis unit 410.
FIG. 9 shows the second sample block 51.
2 and FIG. 10 show error data blocks 513, respectively. In FIG. 10, hatched portions indicate pixels in the second sample block 512 and the reconstructed second sample block 5.
The pixels in 11 do not match (that is, the pixels having the first binary value). On the other hand, in FIG. 10, a portion without oblique lines indicates a pixel in which the pixel in the second sample block 512 matches the pixel in the reconstructed second sample block 511 (that is, the pixel having the second binary value). Represent.

If there are a plurality of first binary values in the error data block, the island coding unit 435 applies, for example, an RCB (Reference Control) to the error data block.
island coding using our own coding technique and supplying the resulting island coded data or coded error data to the data formatting and bit calculation unit 480 via line L49, and also performing island coding. A first island encoding mode signal indicating whether the signal is present or not is supplied to the mode encoding unit 470. However, when there is no first binary pixel in the error data block, only the first island coding mode signal indicating whether or not island coding has been performed by the island coding unit 435 is subjected to mode coding. Section 470. That is, no data is sent to the data formatting and bit calculation unit 480.

Hereinafter, an encoding process of an error data block using the RCB technique will be described with reference to FIG. First, as shown as vertices A to E in FIG. 17, the position of the hatched portion (that is, the first binary pixel) in the error data block is detected.

Thereafter, a plurality of line segments AB, BC, CD, D
A polygon consisting of E and EA is formed. The length of each line segment is defined by the number of pixels forming the line segment. Next, the amount of bits that can represent the longest line segment among the line segments is determined. M representing the length of the longest line segment is 2 ⁿ ≦ m <2
^If the relationship of ^{n + 1} is satisfied, the position of each vertex is encoded using at most n bits.

More specifically, for example, after a reference point RP is obtained by using a raster scanning method, encoding of each vertex is sequentially performed, for example, clockwise, starting from the reference point RP. Referring again to FIG.
, The distance between the reference point RP and the vertex A is encoded using n bits, and the distance between the corresponding vertex and the previous vertex is similarly encoded using the n bits for the next vertex. Encode. Also, if the distance between the previously encoded vertex and the reference point RP is less than m, the minimum number of bits for encoding the current vertex is newly determined, and the current vertex is a new bit that can represent the distance. Is encoded with the determined number of bits.

Mode encoding section 470 generates an inter / intra mode signal indicating whether first enhancement bit calculation section 310 is operating in an inter mode or an intra mode, and then generates a first line selection mode signal. , The second length adjustment mode signal, the first island coding mode signal, and the inter / intra mode signal, and sets the resulting set of coding mode signals on line L4.
8 to the data formatting and bit calculation unit 480.

Next, the data formatting and bit calculation unit 480 determines the received encoding mode signal set, the first and second length adjustment information, the line analysis information for the selected vertical line, and the island encoded data. And supplies the first encoded vertical enhancement layer data to the encoded enhancement layer data selection unit 360 in FIG. 3 through a line L34. Also, the data formatting and bit calculation unit 480
Counts the number of bits of the first encoded vertical enhancement layer data, and transmits the resulting first vertical bit data to the bit comparison unit 350 in FIG.
To supply.

On the other hand, the second enhancement bit calculator 330 has the same structure as the first enhancement bit calculator 310 and performs the same function. Thus, in the case of the intra mode, there is no input to the second enhancement bit calculation unit 330, and in the case of the inter mode, the second
The inputs to the enhancement bit calculation unit 330 are the second sample block input from the subtraction unit 220 through the line L16 and the line L1 from the first line supply unit 235.
7 is a set of vertical insertion lines input through.

As a result, the second enhancement bit calculator 330 supplies the second encoded vertical enhancement layer data to the encoded enhancement layer data selector 360 via the line L32, and also calculates the number of bits of the second encoded vertical enhancement layer data. The calculated second vertical bit data is supplied to the bit comparison unit 350 through the line L31.

In the case of the intra mode, the bit comparing section 35
0 generates a first selection signal based on the first vertical bit data on the line L33 so that the encoding enhancement layer data selection unit 360 selects the first encoding vertical enhancement layer data. However, in the case of the inter mode, the bit comparison unit 350 firstly outputs the line L31
Vertical bit data from line L33 and the first vertical bit data from line L33.
If the first vertical bit data is smaller than the second vertical bit data as compared with the vertical bit data, the bit comparing section 350 sets the encoding enhancement layer data selecting section 36
A second select signal is generated such that 0 selects the first encoded vertical enhancement layer data. When the first vertical bit data is larger than the second vertical bit data,
The bit comparing unit 350 generates a third selection signal so that the encoded enhancement layer data selection unit 360 selects the second encoded vertical enhancement layer data.

Coding enhancement layer data selection unit 3
60 selects the first encoded vertical enhancement layer data input through the line L34 as the encoded vertical enhancement layer data based on the first selection signal,
The signal is supplied to the MUX 250 through the line L18. Conversely, the encoded enhancement layer data selector 360 selects the first encoded vertical enhancement layer data based on the second selection signal and the second encoded vertical enhancement layer data based on the third selection signal. . The coded vertical enhancement layer data selected in this way is used as a signal for notifying whether the coded vertical enhancement layer data has been selected by the coded enhancement layer data selection section 360 through the MUX 25 through the line L18.
0 is supplied.

According to another preferred embodiment of the present invention, the coded enhancement layer data selection unit 360 may perform the selected first coded vertical enhancement layer data and the second coded vertical enhancement layer data in addition to the above operation. Is calculated, and the resulting selection ratio data is supplied to the second enhancement layer coding unit 245 in FIG. 2 through the line L19, for example, in frame units.

According to the second preferred embodiment of the present invention, in the case of the intra mode, the second enhancement layer coding unit 2
45 is the first block from line L11 and line L1
2, the horizontal enhancement layer is encoded based on the second block, and the resulting encoded horizontal enhancement layer data is transmitted via line L21 to MUX2.
Supply 50. In the case of the inter mode, the second
The enhancement layer encoding unit 245 encodes the horizontal enhancement layer based on a set of horizontal insertion lines from the line L20 in addition to the first block from the line L11 and the second block from the line L12. The resulting encoded horizontal enhancement layer data is
X250. Hereinafter, detailed functions of the second enhancement encoding unit 245 will be described by describing points different from the functions of the first enhancement layer encoding unit 230.

First, the first enhancement encoder 23
When the base layer (that is, the first sample block) is input to the interpolation unit 300 within 0 through the line L13, the second layer
A first block of 8 × 16 pixels is provided on a line L in an interpolation unit (not shown) in the enhancement layer coding unit 245.
11 is input. Further, when the second sample block is input to the first enhancement bit calculation unit 310 in the first enhancement encoding unit 230 via the line L16, the first enhancement bit in the second enhancement layer encoding unit 245 The calculation unit (not shown) has a second block composed of 8 × 16 pixels as a line L
12 is input.

The first enhancement layer coding section 2
When a set of vertical insertion lines is input to the second enhancement bit calculator 330 in the line 30 via the line L17, the second enhancement bit calculator (not shown) in the second enhancement layer encoder 245 A set of horizontal insertion lines is input via line L20. According to the embodiment of the present invention, the first enhancement layer coding unit 2
30 is different from the first enhancement layer coding unit 2
Encoding enhancement layer data selection unit 360 in 30
Supplies the selection ratio data to a coded enhancement layer data selector (not shown) in the second enhancement layer encoder 245 via a line L19.

The second enhancement layer coding section 245 supplies the coded horizontal enhancement layer data to the MUX 250 via the line L21 according to the received input data. As described above, the first and second enhancement layer encoders 230 and 245 are similar in structure, but the differences are as follows.

That is, the first enhancement layer coding unit 2
In contrast to the fact that all the lines in the base layer input to 30 and the set of vertical insertion lines in the second sample block are all vertical lines, the first input to the second enhancement layer encoder 245 All the lines in the block and the set of horizontal insertion lines in the second block are all horizontal lines. Therefore, the second enhancement layer encoder 2
All operations performed at 45 are for horizontal lines.

For example, the interpolation method for generating a set of vertical interpolation lines can be applied to obtaining a set of vertical interpolation lines. Turning the set of vertical lines shown in FIG. 15 clockwise by about 90 ° results in a set of vertical lines. Similarly, when the set of horizontal interpolation lines shown in FIG. 16 is rotated clockwise by about 90 °, it becomes a set of vertical interpolation lines. At this time, the horizontal length of the set of horizontal lines is 16 pixels.

Also, the first enhancement layer coding section 2
Encoding enhancement layer data selection unit 360 in 30
, A coded enhancement layer data selector (not shown) in the second enhancement layer coder 245
Has increased the coding efficiency by allocating different numbers of bits based on the selection ratio data input from the coding enhancement layer data selection unit 360 via the line L19. Therefore, for example, a smaller number of bits are assigned to the encoding of the enhancement layer data having the larger selection ratio data among the selection ratios between the first encoded enhancement layer data and the second encoded enhancement layer data.

Although the preferred embodiments of the present invention have been described above, those skilled in the art will be able to make various modifications without departing from the scope of the present invention.

[0089]

Thus, according to the present invention, in coding a video signal, a scalability that gradually increases the resolution of a video is realized, and a binary shape signal is decoded according to a desired video resolution. In addition, it is possible to improve the resolution of a video by reducing errors and bit loss that occur when transmitting a video signal.

[Brief description of the drawings]

FIG. 1 is a block diagram of a binary signal encoding apparatus according to the related art.

FIG. 2 is a block diagram of a binary signal encoding apparatus according to the present invention;

FIG. 3 is a detailed block diagram of a first enhancement layer coding unit in FIG. 2;

FIG. 4 is a detailed block diagram of a first enhancement bit calculator in FIG. 3;

FIG. 5 is a schematic diagram showing a binary alpha block (BAB) and a sample block processed by the binary shape signal encoding apparatus of the present invention.

FIG. 6 is a schematic diagram showing a binary alpha block (BAB) and a sample block processed by the binary shape signal encoding apparatus of the present invention.

FIG. 7 is a schematic diagram showing a binary alpha block (BAB) and a sample block processed by the binary shape signal encoding apparatus of the present invention.

FIG. 8 is a schematic diagram showing a binary alpha block (BAB) and a sample block processed by the binary shape signal encoding apparatus of the present invention.

FIG. 9 is a schematic diagram showing a binary alpha block (BAB) and a sample block processed by the binary shape signal encoding apparatus of the present invention.

FIG. 10 is a schematic diagram showing a binary alpha block (BAB) and a sample block processed by the binary shape signal encoding apparatus of the present invention.

FIG. 11 is a schematic diagram for explaining a vertical interpolation method according to the present invention.

FIG. 12 is a schematic diagram for explaining a vertical interpolation method according to the present invention.

FIG. 13 is a schematic diagram illustrating a vertical interpolation method according to the present invention.

FIG. 14 is a schematic diagram for explaining a vertical interpolation method according to the present invention.

FIG. 15 is a schematic diagram showing a set of vertical lines obtained by the interpolation method of the present invention.

FIG. 16 is a schematic diagram showing a set of vertical interpolation lines obtained by the interpolation method of the present invention.

FIG. 17 is a schematic diagram for explaining an RCB encoding technique used in the present invention.

[Explanation of symbols]

 Reference Signs List 110 Subsampling section 120 Base layer encoding section 130 Enhancement layer encoding section 131 Enhancement layer reconstructing section 205 Horizontal subsampling section 210 Vertical subsampling section 215 Base layer encoding section 220, 225, 430 Subtraction section 235, 240 Line supply unit 230 First enhancement layer coding unit 245 Second enhancement layer coding unit 140, 250 MUX 300 Interpolation unit 310 First enhancement bit calculation unit 330 Second enhancement bit calculation unit 350 Bit comparison unit 360 Coding enhancement layer data Selector 410, 415 Line analyzer 420 Line segment comparator 135, 435 Island encoder 440 Line selector 455, 460 Line adjuster 465 Shape reconstruction unit 470 De coder 480 data formatting and bit calculator

Claims (10)

[Claims] An M × N number (M × N) provided in a video signal
And N are binary shape signal encoding devices for encoding a binary alpha block (BAB) composed of binary pixels each having an even positive integer), wherein each horizontal line of the BAB is encoded. Is sampled, and the B
Horizontal sub-sampling means for generating a first block starting from the first horizontal line located at the top of AB or the second horizontal line located next to the top, and sampling each vertical line of the first block ,
Vertical sub-sampling means for generating, as a base layer, a first sample block starting from a first vertical line that is the leftmost vertical line of the first block or a second vertical line positioned next to the leftmost line, Encoding a first sample block to generate encoded base layer data; and encoding an enhancement layer based on the BAB, the first block, and the first sample block, A binary signal encoding apparatus comprising: enhancement layer encoding means for generating encoded horizontal enhancement layer data and encoded vertical enhancement layer data. 2. The method according to claim 1, wherein each of the video signals includes a plurality of BABs.
2. A set of frames having
2. The binary shape signal encoding device according to item 1. 3. The apparatus according to claim 1, wherein the video signal is a video object plane (VOP) having a plurality of BABs. 4. The apparatus according to claim 2, wherein the video signal has a current frame and a previous frame, and the BAB is in the current frame. 5. When the base layer encoding unit is in an intra mode, the first sample block is encoded using a bitmap-based shape encoding technique,
Encoding means for generating encoded base layer data; and a first frame memory. In the case of the inter mode, the first sample block of the current frame is compared with a corresponding block in a previous frame. Comparing means for selecting a block of a previous frame most similar to the first sample block of the current frame among blocks stored as a first sample block predicted in the first frame memory; Motion vector information determination for obtaining motion vector information represented as a two-dimensional vector composed of a horizontal component and a vertical component, representing a displacement between the first sample block of the previous frame and the predicted first sample block of the previous frame. Means, the first sample block of the current frame and the predicted first sample of the previous frame. And an error data block encoding means for encoding an error data block representing a difference between the motion vector information and the motion vector information and the encoding error data. After generating as data, combining means for determining a reconstructed first sample block in the current frame based on the encoded error data and the predicted first sample block; and 5. The binary shape signal encoding apparatus according to claim 4, further comprising storage means for storing one sample block at a predetermined position in said first frame memory. 6. The enhancement layer encoding means: first subtraction means for subtracting the first sample block from the first block to generate a second sample block; and subtracting the first block from the BAB. And a second subtraction means for generating a second block; and a second frame memory. In the case of the intra mode, the second sample block is stored in the second frame memory, and the In the case of the mode, while storing the second sample block, among the second blocks in the previous frame stored in the second frame memory based on the motion vector information,
Extracting a first extraction block satisfying a first predetermined criterion in which a horizontal and vertical distance between the second block and the first extraction block are horizontal and vertical components of the motion vector, respectively; Vertical line supply means for supplying one extraction block as a set of vertical insertion lines; and, in the case of the intra mode, encoding of a vertical enhancement layer based on the first sample block and the second sample block, The coded vertical enhancement layer data is supplied, and in the case of the inter mode, coding of the vertical enhancement layer is performed based on a set of the first sample block, the second sample block, and the vertical insertion line. First enhancement layer coding means for supplying the generalized vertical enhancement layer data; In the case of the intra mode, the second block is stored in the third frame memory. In the case of the inter mode, the second block is stored. Of the second blocks in the previous frame stored in the third frame memory based on the vector information, the horizontal and vertical distances between the second block and the second fetched block are respectively equal to the motion vector. Horizontal line supply means for extracting a second extraction block satisfying a second predetermined criterion that is twice the horizontal component and one time the vertical component and supplies the second extraction block as a set of horizontal insertion lines; In the case of the intra mode, encoding of a horizontal enhancement layer is performed based on the first block and the second block, and the encoded horizontal enhancement layer is encoded. In the case of the inter mode, encoding of a horizontal enhancement layer is performed based on a set of the first sample block, the second sample block, and the horizontal insertion line. The binary shape signal encoding apparatus according to claim 5, further comprising a second enhancement layer encoding unit that supplies layer data. 7. The first enhancement layer encoding means, comprising: a vertical interpolation means for generating a set of vertical interpolation lines using a predetermined interpolation method; and in the case of the intra mode and the inter mode, Encoding the first vertical enhancement layer based on a set of vertical interpolation lines and the second sample block,
First enhancement layer bit calculating means for generating 1 encoded vertical enhancement layer data, calculating the number of bits of the first encoded vertical enhancement layer data, and generating the calculation result as first vertical bit data; In the case of the intra mode, no operation is performed. In the case of the inter mode, the second enhancement layer is coded based on the set of the vertical insertion lines and the second sample block, and the second coding is performed. A second enhancement layer bit calculating means for supplying the vertical enhancement layer data, calculating the number of bits of the corresponding second encoded vertical enhancement layer data, and generating the calculation result as second vertical bit data; , The first encoded vertical enhancement is performed based on the first vertical bit data. Generating a first selection signal to select the first bit data and the second bit data in the case of the inter mode.
Comparing the bit data, if the first bit data is equal to or less than the second bit data, generate a second selection signal to select the first encoded vertical enhancement layer data; otherwise, generate a second selection signal. Vertical bit comparing means for generating a third selection signal to select the second encoded vertical enhancement layer data; and in the case of the intra mode, the first encoded vertical enhancement signal in response to the first selection signal. Layer data is selected as encoded vertical enhancement layer data, and in the case of the inter mode, the first and second encoded vertical enhancement layer data are encoded according to the second selection signal and the third selection signal. The first and second encoded vertical enhancement layer data are respectively selected as enhancement layer data, and the encoded vertical enhancement layer data is Generating the selected encoded vertical enhancement layer data, which is a signal indicating that the selected encoded vertical enhancement layer data has been selected, and the selected first encoded vertical enhancement layer data and the selected second encoded vertical enhancement layer data. 7. The binary shape signal encoding apparatus according to claim 6, further comprising: first encoding enhancement layer data selecting means for calculating a selection ratio between the enhancement layer data and the corresponding selection ratio. . 8. The image processing apparatus according to claim 1, wherein the second enhancement layer encoding means uses a predetermined interpolation method to generate a set of horizontal interpolation lines, and in the case of the intra mode and the inter mode, Encoding the first horizontal enhancement layer based on a set of horizontal interpolation lines and the second sample block,
A first enhancement layer bit calculating means for generating 1 encoded horizontal enhancement layer data, calculating the number of bits of the first encoded horizontal enhancement layer data, and generating the calculation result as first horizontal bit data; No operation is performed in the case of the intra mode, and in the case of the inter mode, the second enhancement layer is coded based on the set of the horizontal insertion lines and the second sample block, and the second coding is performed. A second enhancement layer bit calculating means for supplying horizontal enhancement layer data, calculating the number of bits of the corresponding second encoded horizontal enhancement layer data, and generating the calculation result as second horizontal bit data; In the case of, the first encoded horizontal encoding is performed based on the first horizontal bit data. The first selection signal generated in order to select the Nsumento layer data, wherein when the inter mode, the said first bit data second
If the first bit data is smaller than the second bit data, a second selection signal is generated to select the first encoded horizontal enhancement layer data. Horizontal bit comparing means for generating a third selection signal to select the second encoded horizontal enhancement layer data; and in the case of the intra mode, the first encoded horizontal enhancement layer data in response to the first selection signal. Selecting data as coded horizontal enhancement layer data, and in the case of the inter mode, coding the first and second coded horizontal enhancement layer data in accordance with the second selection signal and the third selection signal. Layer data, and the first and second coded horizontal enhancement layer data are coded horizontal enhancement methods. Said selected encoded horizontal enhancement layer data is a signal for notifying that it has been selected as the coat layer data occurs,
Calculating a selection ratio between the selected first coded horizontal enhancement layer data and the selected second coded horizontal enhancement layer data, and selecting a first coded enhancement layer data to generate a corresponding selection ratio; The binary shape signal encoding apparatus according to claim 7, comprising means. 9. The first vertical enhancement bit calculating means analyzes the set of vertical interpolation lines to detect the number of line segments in each vertical interpolation line, the length of each line segment, and the positions of the start point and end point. do it,
First line analysis means for generating the detection result as line analysis information of a first sample block; analyzing the set of vertical lines of the second sample block to determine the number of line segments in each vertical line; A second line analyzing means for detecting a length, a start point and an end point position, and generating the detection result as line analysis information of a second sample block; and a fourth frame memory. The line analysis information of the second sample block is stored in the fourth frame memory, and based on the line analysis information of the first and second sample blocks, the number of line segments of each vertical interpolation line and the number of lines in the second sample block are calculated. Is compared with the number of line segments of each corresponding vertical line, and if they match, the first line analysis information for the vertical interpolation line and the vertical line in the second sample block If not, the third and fourth line analysis information for the vertical lines of the second sample block located on the left and right sides of the vertical interpolation line are generated. However, at this time, if the vertical interpolation line is the leftmost interpolation line in the set of the vertical interpolation lines, the leftmost interpolation line in the second sample block stored in the fourth frame memory is used. Generating a vertical line located as third line analysis information for the vertical line instead of the third line analysis information for the leftmost vertical line in the set of vertical interpolation lines; And a line segment comparing means for generating fourth line analysis information, and the first and second line analysis information so as to be the same as the corresponding vertical line in the second sample block. Then, the length of the line segment of each vertical interpolation line is adjusted, and the first line adjustment information indicating the line adjustment state and the second line information indicating whether the adjustment of the length of the line segment has been performed for each line segment. A first line adjustment unit for generating a length adjustment mode signal; and a number of line segments of each of the vertical interpolation lines and a number of line segments of each of the vertical interpolation lines based on the first, third, and fourth line analysis information. Compare with the number of vertical line segments located on the left and right sides, if they do not match, select any one of the left or right vertical lines according to a predetermined selection rule, Generating a line analysis information for the selected vertical line and a first line selection mode signal for notifying that the vertical line has been selected by the predetermined selection rule; Left or right of the vertical interpolation line If the number of line segments of the vertical line located in the same, select a vertical line having the same number of line segments as the vertical interpolation lines, line analysis information for the selected vertical line, A line selection means for generating a second line selection mode signal for notifying that a vertical line has been selected from the right and left vertical lines, and the first line analysis information; The line analysis information for
Based on the first line selection mode signal and the first line analysis information, the length of each vertical interpolation line is adjusted by the same method as the method of the first line adjustment means, and the second A second length adjustment mode signal for generating length adjustment information and a second length adjustment mode signal indicating whether the length adjustment has been performed;
Line adjustment means, the line analysis information for the selected vertical line,
The first line selection mode signal, the second line selection mode signal, the second length adjustment information, the second length adjustment mode signal, the first length adjustment information, and the first length adjustment mode signal Shape reconstruction means for reconstructing the second sample block to form a reconstructed second sample block based on the second sample block, and subtracting the reconstructed second sample block from the second sample block, Subtracting means for supplying the difference as an error data block comprising a first binary value and a second binary value, wherein the first binary value includes the second sample block and the reconstructed second sample. Represents pixels whose binary pixel values do not match each other with the block,
The subtraction means, wherein the second binary value represents a pixel whose binary pixel value matches between the second sample block and the reconstructed second sample block; and 1 binary pixel is 1
If there is one or more of them, the error data block is subjected to island coding, and island coded data as coded error data, and a first indicating whether or not island coding has been performed. Providing an island encoding mode signal, and if no first binary pixel is present in the error data block, the first island encoding indicates that the island encoding was not performed. An island encoding means for supplying only a mode signal; and an inter / intra mode signal for determining whether the first enhancement bit calculation means is currently operating in an intra mode or an inter mode, and the first line selection mode. Signal, the second length adjustment mode signal, the first length adjustment mode signal, the first island coding mode signal. Encoding means for encoding a data signal and the inter / intra mode signal to supply a set of coding mode signals; a set of the coding mode signals, the first length adjustment information, and the second length Formatting the adjustment information, the line analysis information for the selected vertical line, and the island coded data to supply the first coded vertical enhancement layer data, and the number of bits of the first coded vertical enhancement layer data 9. The binary shape signal encoding apparatus according to claim 8, further comprising: a data formatting and bit calculation unit that counts data and supplies the first vertical bit data. 10. The binary shape signal encoding apparatus according to claim 9, wherein M and N are 16 respectively.